Self-assembling nanofiber hydrogels to attenuate epithelial mesenchymal transition in lens
epithelial cells
da Cruz Barros, Raquel Sofia
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2018
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
da Cruz Barros, R. S. (2018). Self-assembling nanofiber hydrogels to attenuate epithelial mesenchymal
transition in lens epithelial cells. University of Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
CHAPTER 3
Epithelial-mesenchymal
transition in lens epithelial cells
attenuated by exposure to
functionalized self-assembling
Submitted to Experimental Eye Research: Barros RC, Gelens E, de Jong MR, Kuijer R, van Kooten TG. Epithelial-mesenchymal transition in lens epithelial cells attenuated by exposure to functionalized self-assembling nanofiber gels (2017)
Functionalized self-assembling nanofiber gels can be used to reduce Epithelial-Mesenchymal Translon in lens epithelial cells associated with Posterior Capsule Opacification (PCO)
The peptide mixture resembling the basement membrane is the best combination for this functionalization
The bioactive hydrogel holds characteristics that can be used to replace the eye lens material
3
Submitted to Experimental Eye Research: Barros RC, Gelens E, de Jong MR, Kuijer R, van Kooten TG. Epithelial-mesenchymal transition in lens epithelial cells attenuated by exposure to functionalized self-assembling nanofiber gels (2017)
HIGHLIGHTS
Functionalized self-assembling nanofiber gels can be used to reduce Epithelial-Mesenchymal Translon in lens epithelial cells associated with Posterior Capsule Opacification (PCO)
The peptide mixture resembling the basement membrane is the best combination for this functionalization
The bioactive hydrogel holds characteristics that can be used to replace the eye lens material
ABSTRACT
3D gels of self-assembling nanofibers based on low molecular weight gelators (LMWG) and functionalized with peptides derived from proteins from the basement membrane: laminin, collagen and fibronectin were used to investigate the effects of matrix composition on lens epithelial cells (LEC) and assess their potential as modulators of epithelial-mesenchymal transition (EMT). LEC were seeded on top of or underneath the gels and were analyzed for metabolic activity, morphology, α-SMA expression and EMT/fibrotic gene expression. LEC seeded on LMWG mixed with peptides had a significant increase in metabolic activity compared with LMWG without peptides. Only LEC seeded on top of the basement membrane (BM) mixture and on Matrigel were spreading and assembled into a monolayer resembling their natural morphology. The expression of α-SMA fibers at the protein level and the mRNA expression of other fibrotic genes were higher on Matrigel. In general, the BM mixture was able to maintain LEC in a lower fibrotic state than the other gels including Matrigel. The data suggest that a cell instructing hydrogel with a proper peptide composition can attenuate EMT.
KEYWORDS
epithelial-mesenchymal transition, lens epithelial cells, self-assembling nanofibers, low molecular weight gelators, fibrosis, hydrogels
INTRODUCTION
Epithelial mesenchymal transition is a process where epithelial cells change their phenotype to mesenchymal-like cells [1]. This process can occur in most of the organs. In several cases, the EMT transformation can become “mature” resulting in a fibrotic tissue [2]. The EMT process is the phenomenological expression of fibrosis in liver, kidney, heart or eye. For none of these diseases there is a specific treatment or cure, which prevents or inhibits the onset of fibrosis.
In the eye lens the fibrotic tissue formed at the posterior part of the capsule bag is named posterior capsule opacification (PCO). This process starts to occur after the removal of the catarogenic nucleus of the lens. The residual lens epithelial cells (LEC) located at the anterior part of the capsule bag sense the inflammatory proteins and autocrine / paracrine factors released during injury and start the EMT transformation [3, 4]. The epithelial cells lose their polarity, which is an important internal regulator for these cells. The polarity is associated with the basement membrane maintaining the interactions with their basal surface [5-7]. The apical surface is more interactive with the extra cellular matrix (ECM) in the fiber cells. At the same time, the epithelial cells also change their phenotype, start to express mesenchymal markers and change their cytoskeleton into a more elongated shape [8]. Lens epithelial cells are relatively small in size when compared with other cells such as mesenchymal cells, and they are organized in a monolayer. These attributes are lost in EMT, the cells get detached from the basement membrane and they are able to migrate to the posterior part of the capsular bag [3]. During the migration these cells are more similar to mesenchymal cells than to epithelial cells and are denominated myofibroblasts. Additional to the high capacity to migrate, these cells also change the actin filaments and start to produce an actin isoform typical of smooth muscle differentiation, alpha-smooth muscle actin (α-SMA) [9]. Then myofibroblast accumulate in the posterior part and the fibrotic tissue starts to be created.
The basement membrane (BM) can be defined as a sandwich between a layer of laminin (the major compound), collagens, fibronectin and proteoglycans and a cell layer [10]. The layers may vary in composition and in the cell connection depending of the tissue of origin [11]. In the native lens it is known that within the basement membrane laminin and collagen are the major constituents and fibronectin is barely detected. The interaction between the cells and the laminin is associated with cell attachment and epithelial differentiation [12, 13]. Different isoforms of laminin have been described, as well as different cell adhesion mediating peptides. IKVAV (Ile-Lys-Val-Ala-Val) derived from the laminin α1 chain is known to promote angiogenesis, cell adhesion and neurite growth [14] and YIGSR (Tyr-Ile-Gly-Ser-Arg) derived from the laminin β1 chain is described as an
inhibitor for tumor cell growth and angiogenesis [15]. Both are short peptide sequences and can be easily used as adhesive sequences by covalently linking them to supporting biomaterials including hydrogel components.
One group of hydrogels comprises the low molecular weight gelators (LMWG) that are produced by self-assembly of low molecular weight molecules. Water molecules are easily entrapped in these networks as well as hydrophobic, hydrophilic and pH sensitive moieties [16]. LMWG based on a 1,3,5-triamide cis,cis-cyclohexane core are synthetic
3
ABSTRACT3D gels of self-assembling nanofibers based on low molecular weight gelators (LMWG) and functionalized with peptides derived from proteins from the basement membrane: laminin, collagen and fibronectin were used to investigate the effects of matrix composition on lens epithelial cells (LEC) and assess their potential as modulators of epithelial-mesenchymal transition (EMT). LEC were seeded on top of or underneath the gels and were analyzed for metabolic activity, morphology, α-SMA expression and EMT/fibrotic gene expression. LEC seeded on LMWG mixed with peptides had a significant increase in metabolic activity compared with LMWG without peptides. Only LEC seeded on top of the basement membrane (BM) mixture and on Matrigel were spreading and assembled into a monolayer resembling their natural morphology. The expression of α-SMA fibers at the protein level and the mRNA expression of other fibrotic genes were higher on Matrigel. In general, the BM mixture was able to maintain LEC in a lower fibrotic state than the other gels including Matrigel. The data suggest that a cell instructing hydrogel with a proper peptide composition can attenuate EMT.
KEYWORDS
epithelial-mesenchymal transition, lens epithelial cells, self-assembling nanofibers, low molecular weight gelators, fibrosis, hydrogels
INTRODUCTION
Epithelial mesenchymal transition is a process where epithelial cells change their phenotype to mesenchymal-like cells [1]. This process can occur in most of the organs. In several cases, the EMT transformation can become “mature” resulting in a fibrotic tissue [2]. The EMT process is the phenomenological expression of fibrosis in liver, kidney, heart or eye. For none of these diseases there is a specific treatment or cure, which prevents or inhibits the onset of fibrosis.
In the eye lens the fibrotic tissue formed at the posterior part of the capsule bag is named posterior capsule opacification (PCO). This process starts to occur after the removal of the catarogenic nucleus of the lens. The residual lens epithelial cells (LEC) located at the anterior part of the capsule bag sense the inflammatory proteins and autocrine / paracrine factors released during injury and start the EMT transformation [3, 4]. The epithelial cells lose their polarity, which is an important internal regulator for these cells. The polarity is associated with the basement membrane maintaining the interactions with their basal surface [5-7]. The apical surface is more interactive with the extra cellular matrix (ECM) in the fiber cells. At the same time, the epithelial cells also change their phenotype, start to express mesenchymal markers and change their cytoskeleton into a more elongated shape [8]. Lens epithelial cells are relatively small in size when compared with other cells such as mesenchymal cells, and they are organized in a monolayer. These attributes are lost in EMT, the cells get detached from the basement membrane and they are able to migrate to the posterior part of the capsular bag [3]. During the migration these cells are more similar to mesenchymal cells than to epithelial cells and are denominated myofibroblasts. Additional to the high capacity to migrate, these cells also change the actin filaments and start to produce an actin isoform typical of smooth muscle differentiation, alpha-smooth muscle actin (α-SMA) [9]. Then myofibroblast accumulate in the posterior part and the fibrotic tissue starts to be created.
The basement membrane (BM) can be defined as a sandwich between a layer of laminin (the major compound), collagens, fibronectin and proteoglycans and a cell layer [10]. The layers may vary in composition and in the cell connection depending of the tissue of origin [11]. In the native lens it is known that within the basement membrane laminin and collagen are the major constituents and fibronectin is barely detected. The interaction between the cells and the laminin is associated with cell attachment and epithelial differentiation [12, 13]. Different isoforms of laminin have been described, as well as different cell adhesion mediating peptides. IKVAV (Ile-Lys-Val-Ala-Val) derived from the laminin α1 chain is known to promote angiogenesis, cell adhesion and neurite growth [14] and YIGSR (Tyr-Ile-Gly-Ser-Arg) derived from the laminin β1 chain is described as an
inhibitor for tumor cell growth and angiogenesis [15]. Both are short peptide sequences and can be easily used as adhesive sequences by covalently linking them to supporting biomaterials including hydrogel components.
One group of hydrogels comprises the low molecular weight gelators (LMWG) that are produced by self-assembly of low molecular weight molecules. Water molecules are easily entrapped in these networks as well as hydrophobic, hydrophilic and pH sensitive moieties [16]. LMWG based on a 1,3,5-triamide cis,cis-cyclohexane core are synthetic
These gels show potential as drug carrier and cell storage depot and can be used both as surface coatings and 3D gels [19]. The 2D nanofibers core of the LMWGs were also studied as a potential biomimetic coating for cell types that are important in the development of fibrosis [20].
Matrigel is commercially available and it has frequently been used as a standard material to study cell interactions with gels. Its composition is not well controlled (Hughes et al. 2010) but it is rich in laminin, collagen IV, nidogen/enactin and proteoglycan, all macromolecules from the basement membrane [21]. It has been referred to as an inductor for cell proliferation [22, 23] although its matrix content cannot be altered to modulate cell responses. Although nowadays Matrigel can be adapted with different (levels of) growth factors, the matrix core is still collected from Engelbroth-Holm-Swarm (EHS) mouse sarcoma cells, which can cause variances between batches [24]. The considerable potential advantage of using synthetic LMWGs is the lack of animal components and the possibility to anchor peptides in a controlled fashion with respect to composition and percentage. These peptides with pre-set ratios and absolute quantities can be used as modulators that direct the cell response.
In this study a two-fold aim was determined. First, a 3D hydrogel model was developed to mimic the situation of LECs in the eye lens. LECs are located on the anterior surface of the lens as a monolayer. To mimic this natural state we used two permutations, one with a layer of LEC on top of the gel and another with a cell layer below a gel. The two situations were compared for the efficiency of LEC in obtaining cues from the matrix. And secondly, cell instructing hydrogelators which build upon a nanofiber-based core activated with peptides present in laminin (LMWG IKVAV+YIGSR) or in a mix of the basement membrane peptides laminin, fibronectin and collagen (LMWG BM) were used to determine whether EMT can be modulated with cues from the matrix in contact with the cells. This aim is based on the knowledge that the dislocation of epithelial cells from the basement membrane is fundamental to the development of EMT [25] and that the role of adhesive proteins is crucial to this linkage. Summarizing, we aim to study the modulation of LEC behavior interacting with self-assembling nanofiber-based LMWG hydrogels equipped with different adhesion peptides. This investigation would open a wider range of possibilities for cell instructing hydrogelators in modifying LEC behavior and decelerating EMT.
MATERIALS AND METHODS
Materials
Lens Epithelial Cells (LEC B3; CRL 11421) were obtained from the American Type Culture Collection (ATCC). Culture medium, Minimum Essential Media containing Earle’s salts (EMEM), fetal bovine serum (FBS), GlutaMax, penicillin, streptomycin and sodium pyruvate were all purchased from Life Technologies, Inc. (Bleiswijk, The Netherlands)). Eight wells Permanox chamber slides were from VWR (Lab-Tek®). Matrigel ® matrix basement membrane with reduced growth factor content was obtained from Corning BD Biosciences. The XTT assay (2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide salt) was used to assess metabolic cell activity (Pan Reac AppliChem (VWR international, Roden, The Netherlands). Triton X-100, bovine serum albumin (BSA), mouse-α-human α-smooth muscle actin antibody, DAPI (4',6-diamidino-2-phenylindole) and phalloidin-TRITC (tetramethylrhodamine isothiocyanate), were from Sigma-Aldrich (Zwijndrecht, The Netherlands). The secondary antibody, FITC (fluorescein isothiocyanate)-labelled goat-α-mouse IgG was from Jackson Immunoresearch Europe (Suffolk, UK). For RNA extraction and PCR we used trizol Pure link RNA Mini kit from Invitrogen (Life Technologies, Rijswijk, The Netherlands), and IQ™SYBR® Green Super Mix from Bio-Rad (Veenendaal, The Netherlands) respectively.
Methods
Low molecular weight gelators production
The used nanofiber-based hydrogel consists on fibers of a low molecular weight hydrogelator (LMWG). In Figure 1 the three mixtures used in this study are shown: LMWG (no peptides), LMWG-laminin (two laminin based peptides, IKVAV and YIGSR) and LMWG-Basement membrane (a combination of laminin peptides (IKVAV+YIGSR), fibronectin peptides (RGDSS+PHSRN) and collagen peptide (DGEA)). The synthesis of LMWG has previously been described by van Bommel et al. [16]. LMWG was functionalized with a maleimide moiety using N-succinimidyl 3-maleimidopropionate (TCI, +95%) and subsequently reacted with cysteine-containing peptides (custom made at Think Peptides, +95% purity) to obtain the building blocks (characterized by 1H-NMR and HPLC-MS) for the different nanogel-peptide composites [26]. The composites were dissolved as 0.15 M solutions in 0.18M HCl in prior to gelification (next section) and sterilized by filtering through a 0.45 µM Whatman filter. 5% Wt hyaluronic acid solution were prepared by storing hyaluronic acid in distilled water. Prior to use the viscous solution was sterilized by filtering through a 0.45 µM Whatman filter.
3
These gels show potential as drug carrier and cell storage depot and can be used both as surface coatings and 3D gels [19]. The 2D nanofibers core of the LMWGs were also studied as a potential biomimetic coating for cell types that are important in the development of fibrosis [20].
Matrigel is commercially available and it has frequently been used as a standard material to study cell interactions with gels. Its composition is not well controlled (Hughes et al. 2010) but it is rich in laminin, collagen IV, nidogen/enactin and proteoglycan, all macromolecules from the basement membrane [21]. It has been referred to as an inductor for cell proliferation [22, 23] although its matrix content cannot be altered to modulate cell responses. Although nowadays Matrigel can be adapted with different (levels of) growth factors, the matrix core is still collected from Engelbroth-Holm-Swarm (EHS) mouse sarcoma cells, which can cause variances between batches [24]. The considerable potential advantage of using synthetic LMWGs is the lack of animal components and the possibility to anchor peptides in a controlled fashion with respect to composition and percentage. These peptides with pre-set ratios and absolute quantities can be used as modulators that direct the cell response.
In this study a two-fold aim was determined. First, a 3D hydrogel model was developed to mimic the situation of LECs in the eye lens. LECs are located on the anterior surface of the lens as a monolayer. To mimic this natural state we used two permutations, one with a layer of LEC on top of the gel and another with a cell layer below a gel. The two situations were compared for the efficiency of LEC in obtaining cues from the matrix. And secondly, cell instructing hydrogelators which build upon a nanofiber-based core activated with peptides present in laminin (LMWG IKVAV+YIGSR) or in a mix of the basement membrane peptides laminin, fibronectin and collagen (LMWG BM) were used to determine whether EMT can be modulated with cues from the matrix in contact with the cells. This aim is based on the knowledge that the dislocation of epithelial cells from the basement membrane is fundamental to the development of EMT [25] and that the role of adhesive proteins is crucial to this linkage. Summarizing, we aim to study the modulation of LEC behavior interacting with self-assembling nanofiber-based LMWG hydrogels equipped with different adhesion peptides. This investigation would open a wider range of possibilities for cell instructing hydrogelators in modifying LEC behavior and decelerating EMT.
MATERIALS AND METHODS
Materials
Lens Epithelial Cells (LEC B3; CRL 11421) were obtained from the American Type Culture Collection (ATCC). Culture medium, Minimum Essential Media containing Earle’s salts (EMEM), fetal bovine serum (FBS), GlutaMax, penicillin, streptomycin and sodium pyruvate were all purchased from Life Technologies, Inc. (Bleiswijk, The Netherlands)). Eight wells Permanox chamber slides were from VWR (Lab-Tek®). Matrigel ® matrix basement membrane with reduced growth factor content was obtained from Corning BD Biosciences. The XTT assay (2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide salt) was used to assess metabolic cell activity (Pan Reac AppliChem (VWR international, Roden, The Netherlands). Triton X-100, bovine serum albumin (BSA), mouse-α-human α-smooth muscle actin antibody, DAPI (4',6-diamidino-2-phenylindole) and phalloidin-TRITC (tetramethylrhodamine isothiocyanate), were from Sigma-Aldrich (Zwijndrecht, The Netherlands). The secondary antibody, FITC (fluorescein isothiocyanate)-labelled goat-α-mouse IgG was from Jackson Immunoresearch Europe (Suffolk, UK). For RNA extraction and PCR we used trizol Pure link RNA Mini kit from Invitrogen (Life Technologies, Rijswijk, The Netherlands), and IQ™SYBR® Green Super Mix from Bio-Rad (Veenendaal, The Netherlands) respectively.
Methods
Low molecular weight gelators production
The used nanofiber-based hydrogel consists on fibers of a low molecular weight hydrogelator (LMWG). In Figure 1 the three mixtures used in this study are shown: LMWG (no peptides), LMWG-laminin (two laminin based peptides, IKVAV and YIGSR) and LMWG-Basement membrane (a combination of laminin peptides (IKVAV+YIGSR), fibronectin peptides (RGDSS+PHSRN) and collagen peptide (DGEA)). The synthesis of LMWG has previously been described by van Bommel et al. [16]. LMWG was functionalized with a maleimide moiety using N-succinimidyl 3-maleimidopropionate (TCI, +95%) and subsequently reacted with cysteine-containing peptides (custom made at Think Peptides, +95% purity) to obtain the building blocks (characterized by 1H-NMR and HPLC-MS) for the different nanogel-peptide composites [26]. The composites were dissolved as 0.15 M solutions in 0.18M HCl in prior to gelification (next section) and sterilized by filtering through a 0.45 µM Whatman filter. 5% Wt hyaluronic acid solution were prepared by storing hyaluronic acid in distilled water. Prior to use the viscous solution was sterilized by filtering through a 0.45 µM Whatman filter.
Figure 1: The range of bioactive low molecular weight gelators without peptides (LMWG) and with different peptides (RGDS [Arg-Gly-Asp-Ser], IKVAV, YIGSR, DGEA [Asp-Gly-Glu-Ala], PHSRN [Pro-His-Ser-Arg-Asn]) coupled. Different combinations and percentage of peptides can be made. This study focus on LMWG alone, LMWG – Laminin (mixture of IKVAV+YIGSR) and LMWG-BM (a combination of IKVAV, YIGSR, RGDSS, PHSRN and DGEA).
Hydrogelators and Matrigel gelification
LMWG, LMWG-IKVAV+YIGSR and LMWG-BM were provided with different percentage of peptides (Table 1) by Nano-FM (NanoFiber Matrices B.V., Groningen, The Netherlands). A 5% wt solution of hyaluronic acid in water was added to medium. To allow the gelification, the previous mixture was added to LMWG, LMWG- IKVAV+YIGSR or LMWG-BM in the proportions 1:1:12.3 (hydrogelator: hyaluronic acid solution: medium). 200 µL of the new solution was pipetted into the wells.
Matrigel was diluted in a cold medium in the proportions 1:1 and quickly, 200 µL of gel was added to the wells. For a strong gelification, all the gels were left at RT for 30 minutes and posteriorly, 250 µL of medium was add above them.
Table 1: Low molecular weight gelators: composition and percentage of peptides
Cell culture
Human lens epithelial cells (LEC-B3) were cultured in EMEM supplemented with 17% FBS, and 1% Pen/Strep, 1% GlutaMAX and 1% of sodium pyruvate in a cell culture incubator at 37°C, 100% humidity and 5%CO2. For each experiment LEC were seeded at 10 000 cells/well (1.25x104/cm2) in 8 wells-Permanox chamber slides. To avoid confluency day 5 was chosen as a limit of culture time with an intermediate time point at day 2.
LEC were seeded in two different conditions: LEC seeded on top of LMWG, LMWG IKVAV+YIGSR or Matrigel, named top model; or LEC seeded on 8 wells-Permanox chamber slides for 12 hours with subsequent addition of the gels described above, named the bottom model (Fig.2). LEC seeded on Permanox chamber slides without addition of any gel were used as controls in both conditions.
3
Figure 1: The range of bioactive low molecular weight gelators without peptides (LMWG) and with different peptides (RGDS [Arg-Gly-Asp-Ser], IKVAV, YIGSR, DGEA [Asp-Gly-Glu-Ala], PHSRN [Pro-His-Ser-Arg-Asn]) coupled. Different combinations and percentage of peptides can be made. This study focus on LMWG alone, LMWG – Laminin (mixture of IKVAV+YIGSR) and LMWG-BM (a combination of IKVAV, YIGSR, RGDSS, PHSRN and DGEA).
Hydrogelators and Matrigel gelification
LMWG, LMWG-IKVAV+YIGSR and LMWG-BM were provided with different percentage of peptides (Table 1) by Nano-FM (NanoFiber Matrices B.V., Groningen, The Netherlands). A 5% wt solution of hyaluronic acid in water was added to medium. To allow the gelification, the previous mixture was added to LMWG, LMWG- IKVAV+YIGSR or LMWG-BM in the proportions 1:1:12.3 (hydrogelator: hyaluronic acid solution: medium). 200 µL of the new solution was pipetted into the wells.
Matrigel was diluted in a cold medium in the proportions 1:1 and quickly, 200 µL of gel was added to the wells. For a strong gelification, all the gels were left at RT for 30 minutes and posteriorly, 250 µL of medium was add above them.
Table 1: Low molecular weight gelators: composition and percentage of peptides
Cell culture
Human lens epithelial cells (LEC-B3) were cultured in EMEM supplemented with 17% FBS, and 1% Pen/Strep, 1% GlutaMAX and 1% of sodium pyruvate in a cell culture incubator at 37°C, 100% humidity and 5%CO2. For each experiment LEC were seeded at 10 000 cells/well (1.25x104/cm2) in 8 wells-Permanox chamber slides. To avoid confluency day 5 was chosen as a limit of culture time with an intermediate time point at day 2.
LEC were seeded in two different conditions: LEC seeded on top of LMWG, LMWG IKVAV+YIGSR or Matrigel, named top model; or LEC seeded on 8 wells-Permanox chamber slides for 12 hours with subsequent addition of the gels described above, named the bottom model (Fig.2). LEC seeded on Permanox chamber slides without addition of any gel were used as controls in both conditions.
Figure 2: Illustration for LEC seeded in top (A) and bottom condition (B). In the top condition, LEC were in direct contact with hydrogelators or Matrigel and medium. In bottom condition, LEC were in contact with the hydrogelators or Matrigel and Permanox.
Immunohistochemistry
Four samples for each gel material and each day were fixed with 3.7% paraformaldehyde during 30 minutes at RT. Then they were washed in PBS followed by 5 minutes with 0.5% triton X-100 for membrane permeabilization. The blocking solution, 5% BSA in PBS, was incubated during 30 minutes at RT. After the primary antibody α-SMA mouse-anti-human was diluted at 1:100 in 1%BSA (1% PBSA) in PBS and added to the samples for 1 hour at RT. Cells were washed 3 times in 1% PBSA and the secondary antibodies were incubated for 2 hour at RT. The nuclei were stained by DAPI (2 μg/mL f.c.), the actin filaments were stained by TRITC-labelled phalloidin (2 μg/mL) and the α-SMA were stained by FITC- goat-anti-mouse (2 μg/mL). Samples were analyzed by confocal laser microscopy (LEICA TCS SP2) with a UV laser, Argon laser and Helium laser. Images were acquiring through a 40x immersion objective lens (NA 0.80) with a resolution of 1024x1024 pixels. The experiment was repeated three times.
Cell metabolic assay
Tetrazolium dye, XTT (2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide salt), is reduced to an orange formazan in presence of the products from the mitochondrial oxireductase enzymes. The color of the formazan was measured by absorbance and it is directly proportional to the cell metabolic activity. Therefore the XTT assay was used to quantify the changes in cell metabolic activity between day 2 and day 5. XTT and activation solution were directly added at each well in a proportion 2:1 (medium : XTT). Four samples per gel, one sample of cells and gels alone were incubated at 37°C and 5% CO2 for 4 hours. Subsequently gels were destroyed by pipetting and absorbance was measured at 485 nm and, as reference, at 690 nm using a Fluostar OPTIMA plate reader. Three repetitions of the experiment were performed.
Real time PCR
mRNA of six samples per condition was extracted using Trizol® reagent with the mRNA of six samples per condition was extracted using Trizol® reagent with the PureLink™ RNA mini kit, according to the instruction of the manufacturer. For the lyses of the cells, trizol (1mL per 10 cm2) was added to each well and the bottoms of all samples were scraped. Lysates were incubated at RT for 5 minutes after which chloroform was added. After vigorously shaking, samples were centrifuge at 12 000xg for 15 minutes. The upper aqueous phase, which contained the RNA, was transferred to RNase-free tube and an equal volume of 70% ethanol was added. This solution was mixed and then centrifuged in a fresh spin cartridge tube, which was subsequently washed with washing buffers I and II. The RNA was eluted from the filter in 40µL of RNase-free water and quantified by Nanodrop spectrophotometer.
Genomic DNA was removed from the RNA samples using the DNA-freeTM kit (Ambion), and cDNA was synthesized using the iScript cDNA synthesis kit (BioRad). Quantitative real time PCR for cells cultured for 5 days was performed in duplicate in 384 wells plates using IQ™SYBR® Green Super Mix on a CFX 384™ Real Time PCR System (BioRad). The amplification cycles were: one cycle of 95°C for 3 minutes and 95°C for 10 seconds, followed of 39 cycles at 95°C for 2 seconds, 60°C (for all primers except those for α-SMA) or 57.5°C (for α-SMA primers) for 30 seconds, 60°C for 5 seconds and finally 90°C for 5 seconds. 18S RNA was used as a reference working at 57.5°C and 60°C. Three repetitions of the experiment were performed. Primers were designed using the online Primer-3 software at the NCBI website and were synthesized by Biolegio (Nijmegen, The Netherlands) (Table 2). Genomic data were normalized with samples from LEC B3 cells grown on Permanox for 5 days using the 2-ΔΔCt method [27]. Briefly, an internal normalization with the reference gene (18s) was done to correct the different amounts of RNA per sample (ΔCt). And a posterior normalization for LEC B3 cells without gels was used to calculate the (ΔΔCt). The amount of the target is then calculated by 2-ΔΔCt.
3
Figure 2: Illustration for LEC seeded in top (A) and bottom condition (B). In the top condition, LEC were in direct contact with hydrogelators or Matrigel and medium. In bottom condition, LEC were in contact with the hydrogelators or Matrigel and Permanox.
Immunohistochemistry
Four samples for each gel material and each day were fixed with 3.7% paraformaldehyde during 30 minutes at RT. Then they were washed in PBS followed by 5 minutes with 0.5% triton X-100 for membrane permeabilization. The blocking solution, 5% BSA in PBS, was incubated during 30 minutes at RT. After the primary antibody α-SMA mouse-anti-human was diluted at 1:100 in 1%BSA (1% PBSA) in PBS and added to the samples for 1 hour at RT. Cells were washed 3 times in 1% PBSA and the secondary antibodies were incubated for 2 hour at RT. The nuclei were stained by DAPI (2 μg/mL f.c.), the actin filaments were stained by TRITC-labelled phalloidin (2 μg/mL) and the α-SMA were stained by FITC- goat-anti-mouse (2 μg/mL). Samples were analyzed by confocal laser microscopy (LEICA TCS SP2) with a UV laser, Argon laser and Helium laser. Images were acquiring through a 40x immersion objective lens (NA 0.80) with a resolution of 1024x1024 pixels. The experiment was repeated three times.
Cell metabolic assay
Tetrazolium dye, XTT (2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide salt), is reduced to an orange formazan in presence of the products from the mitochondrial oxireductase enzymes. The color of the formazan was measured by absorbance and it is directly proportional to the cell metabolic activity. Therefore the XTT assay was used to quantify the changes in cell metabolic activity between day 2 and day 5. XTT and activation solution were directly added at each well in a proportion 2:1 (medium : XTT). Four samples per gel, one sample of cells and gels alone were incubated at 37°C and 5% CO2 for 4 hours. Subsequently gels were destroyed by pipetting and absorbance was measured at 485 nm and, as reference, at 690 nm using a Fluostar OPTIMA plate reader. Three repetitions of the experiment were performed.
Real time PCR
mRNA of six samples per condition was extracted using Trizol® reagent with the mRNA of six samples per condition was extracted using Trizol® reagent with the PureLink™ RNA mini kit, according to the instruction of the manufacturer. For the lyses of the cells, trizol (1mL per 10 cm2) was added to each well and the bottoms of all samples were scraped. Lysates were incubated at RT for 5 minutes after which chloroform was added. After vigorously shaking, samples were centrifuge at 12 000xg for 15 minutes. The upper aqueous phase, which contained the RNA, was transferred to RNase-free tube and an equal volume of 70% ethanol was added. This solution was mixed and then centrifuged in a fresh spin cartridge tube, which was subsequently washed with washing buffers I and II. The RNA was eluted from the filter in 40µL of RNase-free water and quantified by Nanodrop spectrophotometer.
Genomic DNA was removed from the RNA samples using the DNA-freeTM kit (Ambion), and cDNA was synthesized using the iScript cDNA synthesis kit (BioRad). Quantitative real time PCR for cells cultured for 5 days was performed in duplicate in 384 wells plates using IQ™SYBR® Green Super Mix on a CFX 384™ Real Time PCR System (BioRad). The amplification cycles were: one cycle of 95°C for 3 minutes and 95°C for 10 seconds, followed of 39 cycles at 95°C for 2 seconds, 60°C (for all primers except those for α-SMA) or 57.5°C (for α-SMA primers) for 30 seconds, 60°C for 5 seconds and finally 90°C for 5 seconds. 18S RNA was used as a reference working at 57.5°C and 60°C. Three repetitions of the experiment were performed. Primers were designed using the online Primer-3 software at the NCBI website and were synthesized by Biolegio (Nijmegen, The Netherlands) (Table 2). Genomic data were normalized with samples from LEC B3 cells grown on Permanox for 5 days using the 2-ΔΔCt method [27]. Briefly, an internal normalization with the reference gene (18s) was done to correct the different amounts of RNA per sample (ΔCt). And a posterior normalization for LEC B3 cells without gels was used to calculate the (ΔΔCt). The amount of the target is then calculated by 2-ΔΔCt.
Statistical analysis
For XTT data one way ANOVA followed by all Pairwise Multiple Comparison Procedures (Bonferroni t-test) was applied in order to compare the difference between hydrogels and between controls.
For analysis of RT-PCR data comparing all the gels with each other a one way ANOVA repeated measurements was performed using Holm-Sidak analysis.
Values were significantly different between each other when p<0.05. All the statistical analysis was performed using SigmaPlot 11.0.
RESULTS
Immunohistochemistry
One of the best characterized markers to detect EMT is α-SMA. The expression of this protein was analyzed by confocal microscopy on day 2 and day 5 of culture on top of and on bottom of the different gels (Fig. 3). LECs seeded on Permanox were used as control. At day 0 of culture, the cells used for the top condition showed an incomplete cytoskeleton. LEC were still creating connections to the gel components (Fig.3A). At day 0, for the bottom condition, (12h of seeding) LEC were completely attached to the Permanox (Fig.3D). The difference between controls on top and bottom remained that in the bottom model all the cells were seeded 12 hours prior to the deposition of the gels.
On day 2 of the top model, cells appeared as clusters in all the gels. These clusters were small but in a higher number on LMWG IKVAV+YIGSR and BM. On day 5 the clusters were transformed into a layer of cells in all gels with exception for LMWG alone, where the size of the clusters increased but their number decreased. This phenomenon demonstrated that LMWG by itself was unsuitable to stimulate proper cell adhesion. From day 2 to day 5 in the LMWG IKVAV+YIGSR the presence of some clusters of cells encased in a layer of cells was observed (Fig.3B). These cells showed a clear proliferation with production of α-SMA mainly in their extensions. On the other hand, cells on LMWG BM showed a total transformation from clusters to a full layer of cells with few signs of α-SMA production. Cells on Matrigel also appeared in a layer of cells on day 5, however a much higher quantity of α-SMA was detected (Fig. 3C). At day 5 the increase in size of the LEC´s nuclei (DAPI staining) and the presence of stress fibers (phalloidin-TRITC staining) were observed on LMWG BM and Matrigel compared with the control.
In the bottom model (Fig. 3E), LEC grew into a layer for 12 hours before the addition of the gels. Although in all the conditions cells maintained their layers, it was established that on day 2 the nuclei of the cells were larger under the gels than in the control. Furthermore α-SMA was present in cells underneath all the gels. With LMWG IKVAV+YIGSR, α-SMA
diffuse staining near the nucleus. The quantity of this protein decreased on day 5 for LMWG IKVAV+YIGSR and LMWG BM but not for Matrigel. This last gel kept the accumulation of α-SMA near the nucleus of the cells (Fig. 3F). In all the gels, with exception of LMWG alone, the size of the nuclei decreased from day 2 to day 5 and the cells became more similar to the control.
In general LMWG BM was the gel that exhibited the smallest quantity of α-SMA and Matrigel showed the highest quantity of α-SMA in both gel-cell permutations.
3
Statistical analysis
For XTT data one way ANOVA followed by all Pairwise Multiple Comparison Procedures (Bonferroni t-test) was applied in order to compare the difference between hydrogels and between controls.
For analysis of RT-PCR data comparing all the gels with each other a one way ANOVA repeated measurements was performed using Holm-Sidak analysis.
Values were significantly different between each other when p<0.05. All the statistical analysis was performed using SigmaPlot 11.0.
RESULTS
Immunohistochemistry
One of the best characterized markers to detect EMT is α-SMA. The expression of this protein was analyzed by confocal microscopy on day 2 and day 5 of culture on top of and on bottom of the different gels (Fig. 3). LECs seeded on Permanox were used as control. At day 0 of culture, the cells used for the top condition showed an incomplete cytoskeleton. LEC were still creating connections to the gel components (Fig.3A). At day 0, for the bottom condition, (12h of seeding) LEC were completely attached to the Permanox (Fig.3D). The difference between controls on top and bottom remained that in the bottom model all the cells were seeded 12 hours prior to the deposition of the gels.
On day 2 of the top model, cells appeared as clusters in all the gels. These clusters were small but in a higher number on LMWG IKVAV+YIGSR and BM. On day 5 the clusters were transformed into a layer of cells in all gels with exception for LMWG alone, where the size of the clusters increased but their number decreased. This phenomenon demonstrated that LMWG by itself was unsuitable to stimulate proper cell adhesion. From day 2 to day 5 in the LMWG IKVAV+YIGSR the presence of some clusters of cells encased in a layer of cells was observed (Fig.3B). These cells showed a clear proliferation with production of α-SMA mainly in their extensions. On the other hand, cells on LMWG BM showed a total transformation from clusters to a full layer of cells with few signs of α-SMA production. Cells on Matrigel also appeared in a layer of cells on day 5, however a much higher quantity of α-SMA was detected (Fig. 3C). At day 5 the increase in size of the LEC´s nuclei (DAPI staining) and the presence of stress fibers (phalloidin-TRITC staining) were observed on LMWG BM and Matrigel compared with the control.
In the bottom model (Fig. 3E), LEC grew into a layer for 12 hours before the addition of the gels. Although in all the conditions cells maintained their layers, it was established that on day 2 the nuclei of the cells were larger under the gels than in the control. Furthermore α-SMA was present in cells underneath all the gels. With LMWG IKVAV+YIGSR, α-SMA
diffuse staining near the nucleus. The quantity of this protein decreased on day 5 for LMWG IKVAV+YIGSR and LMWG BM but not for Matrigel. This last gel kept the accumulation of α-SMA near the nucleus of the cells (Fig. 3F). In all the gels, with exception of LMWG alone, the size of the nuclei decreased from day 2 to day 5 and the cells became more similar to the control.
In general LMWG BM was the gel that exhibited the smallest quantity of α-SMA and Matrigel showed the highest quantity of α-SMA in both gel-cell permutations.
Figure 3: Analysis, by confocal microscopy, of LEC seeded on Permanox chamber slides at day 0 of culture (A, D), LEC seeded on top (B) or on bottom (E) of the LMWG, LMWG IKVAV+YIGSR, LMWG BM and Matrigel. Same cells seeded on Permanox chamber slides were used as controls. Details of the nucleus and α- SMA are also shown for top (C) and bottom (F) conditions. Blue: DAPI (nucleus), red: TRITC-phalloidin (actin fibers), green: FITC- α- SMA. Resolution 1024x1024 pixels corresponding to 375 x 375 µm.
Cell metabolic activity
Both conditions of cell seeding, top and bottom, were analyzed with XTT assays after 2 and 5 days of culture with the different high molecular weight gelators (LMWG alone, LMWG IKVAV+YIGSR or BM) and Matrigel. LEC without hydrogels were also measured as control for both conditions. The changes in absorbance between day 5 and day 2 (Δt) for LEC seeded on top of the hydrogels are showed on figure 4A. Cells on LMWG alone did not show signals of metabolic activity changes between these time points. In combination with the peptides these values increased by 41.7% and 51.7% for IKVAV+YIGSR and BM respectively. Matrigel had the highest metabolic activity when compared with the other materials, but still significantly lower than LEC without the presence of hydrogels (Fig.4 a1).
When LEC were seeded on the bottom of the Permanox with the gels on top of them they sense these two different surfaces, Permanox – a cell-adhesion supporting surface – from the bottom side and gels above. In figure 4B it can be seen that cells exhibited a low metabolic activity for all gel materials with no significant differences in metabolic activity between them (Fig.4 b1).
3
Figure 3: Analysis, by confocal microscopy, of LEC seeded on Permanox chamber slides at day 0 of culture (A, D), LEC seeded on top (B) or on bottom (E) of the LMWG, LMWG IKVAV+YIGSR, LMWG BM and Matrigel. Same cells seeded on Permanox chamber slides were used as controls. Details of the nucleus and α- SMA are also shown for top (C) and bottom (F) conditions. Blue: DAPI (nucleus), red: TRITC-phalloidin (actin fibers), green: FITC- α- SMA. Resolution 1024x1024 pixels corresponding to 375 x 375 µm.
Cell metabolic activity
Both conditions of cell seeding, top and bottom, were analyzed with XTT assays after 2 and 5 days of culture with the different high molecular weight gelators (LMWG alone, LMWG IKVAV+YIGSR or BM) and Matrigel. LEC without hydrogels were also measured as control for both conditions. The changes in absorbance between day 5 and day 2 (Δt) for LEC seeded on top of the hydrogels are showed on figure 4A. Cells on LMWG alone did not show signals of metabolic activity changes between these time points. In combination with the peptides these values increased by 41.7% and 51.7% for IKVAV+YIGSR and BM respectively. Matrigel had the highest metabolic activity when compared with the other materials, but still significantly lower than LEC without the presence of hydrogels (Fig.4 a1).
When LEC were seeded on the bottom of the Permanox with the gels on top of them they sense these two different surfaces, Permanox – a cell-adhesion supporting surface – from the bottom side and gels above. In figure 4B it can be seen that cells exhibited a low metabolic activity for all gel materials with no significant differences in metabolic activity between them (Fig.4 b1).
Figure 4: Variance between day 5 and day 2 (Δt) on cell proliferation measurements by XTT assays. Data is showed using lens epithelial cells (LEC) as 100% of absorbance. A) Lens epithelial cells (LEC) seeded on top of the LMWG, LMWG IKVAV+YIGSR, LMWG BM, Matrigel or LEC without contact with any hydrogel (set as 100). a.1) Details from the statistical analysis for the top model. B) Lens epithelial cells in bottom model, meaning cells seeded below the LMWG, LMWG IKVAV YIGSR, LMWG BM, Matrigel or LEC without contact with any hydrogel. b.1) Details from the statistical analysis for the bottom model. Analysis performed by One-way ANOVA. Significant difference for p<0.05. Error bars depict standard deviations.
Gene expression analysis by rt-PCR
Expression of fibrosis-associated genes (the TGF-βs receptors -ALK2, ALK5, type III and VI collagen, TGF-β1, -β2 and α-SMA) were measured by real-time rt-PCR. Normalized
fold expression was related to LEC cultured on Permanox during 5 days. (set at 1 axis, Fig. 5).
For the top model, cells on LMWG BM expressed a lower level of ALK2 compared with all the other materials. The expression of ALK5 had its lowest level on cells in contact with Matrigel followed by LMWG BM and LMWG. For collagen type III, LEC did not show significant differences between the different hydrogels, although all the values were downregulated compared with LEC without contact with hydrogelators. Collagen type VI
was upregulated in cells on LMWG followed by LMWG IKVAV+YIGSR. The lowest transcription of this gene was on LEC with LMWG BM and Matrigel. The cells on Matrigel
and LMWG IKVAV+YIGSR gave a similar large regulation on of TGF-β1. However, this
gene did not show significant differences between the different gels. Similarly, TGF-β2
appeared downregulated in all the hydrogelators without significant differences between them. The expression of the α-SMA mRNA had similar values in cells on Matrigel, LMWG BM and LMWG, but was downregulated compared with LEC on Permanox (Fig. 5 and S2). For the bottom model, ALK2 appeared downregulated in all cells in contact with the materials. ALK5 was upregulated in cells in contact with LMWG and its expression was significantly lower for LMWG IKVAV YIGSR, BM and Matrigel. The expression of collagen
type III on cells underneath LMWG BM was significantly lower than Matrigel. Cells
underneath the hydrogelators demonstrated a lower expression of this gene when compared with LEC without contact with any gel. The same pattern was detected for
collagen type VI. TGF-β1 showed a large downregulation for LEC in all materials. LEC
under LMWG BM showed a significant downregulation compared with Matrigel. TGF-β2 in
cells under hydrogelators did not show significant differences between materials. For α-SMA all LEC under materials were downregulated compared with LEC without contact with gels. Cells in contact with the hydrogelators, LMWG, LMWG BM and LMWG IKVAV YIGSR expressed significant lower levels of this gene relatively to Matrigel (Fig. 5 and Fig. S2).
3
Figure 4: Variance between day 5 and day 2 (Δt) on cell proliferation measurements by XTT assays. Data is showed using lens epithelial cells (LEC) as 100% of absorbance. A) Lens epithelial cells (LEC) seeded on top of the LMWG, LMWG IKVAV+YIGSR, LMWG BM, Matrigel or LEC without contact with any hydrogel (set as 100). a.1) Details from the statistical analysis for the top model. B) Lens epithelial cells in bottom model, meaning cells seeded below the LMWG, LMWG IKVAV YIGSR, LMWG BM, Matrigel or LEC without contact with any hydrogel. b.1) Details from the statistical analysis for the bottom model. Analysis performed by One-way ANOVA. Significant difference for p<0.05. Error bars depict standard deviations.
Gene expression analysis by rt-PCR
Expression of fibrosis-associated genes (the TGF-βs receptors -ALK2, ALK5, type III and VI collagen, TGF-β1, -β2 and α-SMA) were measured by real-time rt-PCR. Normalized
fold expression was related to LEC cultured on Permanox during 5 days. (set at 1 axis, Fig. 5).
For the top model, cells on LMWG BM expressed a lower level of ALK2 compared with all the other materials. The expression of ALK5 had its lowest level on cells in contact with Matrigel followed by LMWG BM and LMWG. For collagen type III, LEC did not show significant differences between the different hydrogels, although all the values were downregulated compared with LEC without contact with hydrogelators. Collagen type VI
was upregulated in cells on LMWG followed by LMWG IKVAV+YIGSR. The lowest transcription of this gene was on LEC with LMWG BM and Matrigel. The cells on Matrigel
and LMWG IKVAV+YIGSR gave a similar large regulation on of TGF-β1. However, this
gene did not show significant differences between the different gels. Similarly, TGF-β2
appeared downregulated in all the hydrogelators without significant differences between them. The expression of the α-SMA mRNA had similar values in cells on Matrigel, LMWG BM and LMWG, but was downregulated compared with LEC on Permanox (Fig. 5 and S2). For the bottom model, ALK2 appeared downregulated in all cells in contact with the materials. ALK5 was upregulated in cells in contact with LMWG and its expression was significantly lower for LMWG IKVAV YIGSR, BM and Matrigel. The expression of collagen
type III on cells underneath LMWG BM was significantly lower than Matrigel. Cells
underneath the hydrogelators demonstrated a lower expression of this gene when compared with LEC without contact with any gel. The same pattern was detected for
collagen type VI. TGF-β1 showed a large downregulation for LEC in all materials. LEC
under LMWG BM showed a significant downregulation compared with Matrigel. TGF-β2 in
cells under hydrogelators did not show significant differences between materials. For α-SMA all LEC under materials were downregulated compared with LEC without contact with gels. Cells in contact with the hydrogelators, LMWG, LMWG BM and LMWG IKVAV YIGSR expressed significant lower levels of this gene relatively to Matrigel (Fig. 5 and Fig. S2).
Figure 5: mRNA levels of ALK2 (A), ALK5 (B), collagen type III (C), collagen type VI (D), TGF-β1 (E), TGF-β2
(F) and α-SMA (G) analyzed by real time PCR on LEC seeded for 5 days on top (blue color) or on bottom (orange color) of LMWG, LMWG YIGSR+IKVAV, LMWG BM and Matrigel. The control – LEC without contact with any gel- for both conditions is set at 1. Statistical analysis is given in the supplementary information (Fig.S2).
DISCUSSION
In this study a 3D hydrogel model was exploited to analyze LEC behavior in contact with hydrogelators alone or hydrogelators functionalized with the peptides IKVAV and YIGSR or with a mix of peptides from the basement membrane (RGDS, IKVAV, YIGSR, DGEA and PHSRN) denominated BM mix. The aim was to manipulate cell behavior towards decreasing epithelial mesenchymal transition (EMT). The LMWG core was chosen based on studying the effects of nanofiber coatings on control cell types using different molecular building blocks [20]. From this study a self-assembling nanofiber core motif was chosen as a potential material to control EMT. EMT occurs in eye lenses after cataract surgery and is responsible for a fibrotic response indicated as posterior capsular opacification (PCO). Many studies addressed potential strategies to prevent or reduce EMT in the eye lens, for example by interfering with important pathways induced by TGF-β signaling [28], including Smad [29], using miRNA [30] or targeting MMPs [31]. The material [32] and the shape [33] of the intra-ocular lens used to replace the interior of the lens also has been a focus of attention. Here we aimed to develop a model to study and manipulate lens epithelial cells (LEC) and at the same time analyze LEC behavior towards the reduction of EMT using cell instructing hydrogelators equipped with different adhesion peptide motifs.
The eye lens has already been referred to as a model to study organ fibrosis due to its pronounced and well-documented PCO response [34]. In fact, the eye lens poses a simplified model to analyze different fibrosis-associated parameters in a controlled ex vivo environment. The lens of the eye consists of a capsular bag lined along the anterior inside with a single cell type in a monolayer setting, the lens epithelial cells. There is no innervation or vascular system present. This is why the use of LEC in an in vitro model can be more easily related to the in vivo situation than in other organ models. Still, the lens environment is drastically modified after a cataract surgery. The inflammatory response can lead to several changes including the increase of proteins surrounding the lens [35]. This new environment seems to be crucial for the development of EMT and subsequent PCO. In our model the use of serum proteins in the medium was chosen to represent this environment.
LEC are characterized with well-defined apical and basal surfaces [7, 36]. The apical surfaces of the cells are touching the nucleus of the lens and the basal surfaces are attached to the capsular bag through the basal lamina [37]. The interaction of these cells with biomaterials can create a different cell response. For this reason, this study addressed the contact with different hydrogelators on top of the cells and underneath the cells. The interaction with the hydrogelators located above and below the cells at the same time (sandwich model) was excluded from this study due to cell death on all the materials already after two days of incubation. We speculated that this behavior was due to a total and lasting loss of LEC polarity.
LEC showed different behavior when in contact with gels from the top or from the bottom. Cells displayed a higher metabolic activity when seeded on top of the gels than
3
Figure 5: mRNA levels of ALK2 (A), ALK5 (B), collagen type III (C), collagen type VI (D), TGF-β1 (E), TGF-β2
(F) and α-SMA (G) analyzed by real time PCR on LEC seeded for 5 days on top (blue color) or on bottom (orange color) of LMWG, LMWG YIGSR+IKVAV, LMWG BM and Matrigel. The control – LEC without contact with any gel- for both conditions is set at 1. Statistical analysis is given in the supplementary information (Fig.S2).
DISCUSSION
In this study a 3D hydrogel model was exploited to analyze LEC behavior in contact with hydrogelators alone or hydrogelators functionalized with the peptides IKVAV and YIGSR or with a mix of peptides from the basement membrane (RGDS, IKVAV, YIGSR, DGEA and PHSRN) denominated BM mix. The aim was to manipulate cell behavior towards decreasing epithelial mesenchymal transition (EMT). The LMWG core was chosen based on studying the effects of nanofiber coatings on control cell types using different molecular building blocks [20]. From this study a self-assembling nanofiber core motif was chosen as a potential material to control EMT. EMT occurs in eye lenses after cataract surgery and is responsible for a fibrotic response indicated as posterior capsular opacification (PCO). Many studies addressed potential strategies to prevent or reduce EMT in the eye lens, for example by interfering with important pathways induced by TGF-β signaling [28], including Smad [29], using miRNA [30] or targeting MMPs [31]. The material [32] and the shape [33] of the intra-ocular lens used to replace the interior of the lens also has been a focus of attention. Here we aimed to develop a model to study and manipulate lens epithelial cells (LEC) and at the same time analyze LEC behavior towards the reduction of EMT using cell instructing hydrogelators equipped with different adhesion peptide motifs.
The eye lens has already been referred to as a model to study organ fibrosis due to its pronounced and well-documented PCO response [34]. In fact, the eye lens poses a simplified model to analyze different fibrosis-associated parameters in a controlled ex vivo environment. The lens of the eye consists of a capsular bag lined along the anterior inside with a single cell type in a monolayer setting, the lens epithelial cells. There is no innervation or vascular system present. This is why the use of LEC in an in vitro model can be more easily related to the in vivo situation than in other organ models. Still, the lens environment is drastically modified after a cataract surgery. The inflammatory response can lead to several changes including the increase of proteins surrounding the lens [35]. This new environment seems to be crucial for the development of EMT and subsequent PCO. In our model the use of serum proteins in the medium was chosen to represent this environment.
LEC are characterized with well-defined apical and basal surfaces [7, 36]. The apical surfaces of the cells are touching the nucleus of the lens and the basal surfaces are attached to the capsular bag through the basal lamina [37]. The interaction of these cells with biomaterials can create a different cell response. For this reason, this study addressed the contact with different hydrogelators on top of the cells and underneath the cells. The interaction with the hydrogelators located above and below the cells at the same time (sandwich model) was excluded from this study due to cell death on all the materials already after two days of incubation. We speculated that this behavior was due to a total and lasting loss of LEC polarity.
LEC showed different behavior when in contact with gels from the top or from the bottom. Cells displayed a higher metabolic activity when seeded on top of the gels than
morphology must be taken into consideration. For Matrigel the top model is the seeding geometry advised by the supplier. Therefore, the high values for Δt (difference between cell metabolic activity on day 5 and day 2) were expected. A low metabolic activity of LEC on top of LMWG (without peptides) was synonym for quiescent cells or apoptotic cells, present in cell clusters. This indicates that LMWG behaves as a repellent gel that does not readily support cell adhesion despite the presence of a three-dimensional nanofiber network. The addition of peptides to the LMWG yielded an increase in cell metabolic activity mainly for the mixture of peptides that resembles the basement membrane most, the LMWG BM. This was associated with a transition of cell clusters to a more natural-like monolayer of cells. The basal lamina is the natural surface for epithelial cell attachment [38]. However, when the cells were seeded underneath the gels, the metabolic activity ratio was independent of the gel composition with a clear cell layer formation from day 2 to day 5 of culture. This can be explained by the favorable interactions between Permanox and LEC when compared with LEC-gel interactions. In this case, the low metabolic activity cannot be referred to as stress or quiescence but rather as a passive stage where the cells have reached a comfortable cell attachment with corresponding low metabolic activity. Permanox was chosen for this study after several experiments with other cell culture bottoms such as type I collagen and fibronectin coatings and glass (data not shown). In all these cases LEC detached from these materials as an intact layer of cells after addition of the gel components.
Besides the increase of α-SMA during the ongoing EMT process, the shape of the cells also dramatically changes. During EMT the small and hexagonal epithelial cells transform into much larger myofibroblasts that are more similar to mesenchymal cells [8, 39]. When the EMT process has “matured” and cells end up in a fibrotic stage, cells will become more similar to fibroblasts. In general, different stages of EMT can be related with a different marker distribution. Nagamoto et al. reported that mammalian LEC start to express α-SMA almost immediately after culture of the primary cells, which by the way is usually done on stiff materials such as tissue culture plastics [9]. This was also visible in our control cultures where a slight increase of this protein could be observed. Garcia et al. mentioned that it is more appropriate to analyze α-SMA with both qualitative and quantitative tools rather than only detection of the protein. In that study LEC explants from bovine, rabbit and human lenses expressed α-SMA from the start of the culture in the equatorial zones of the lens but not in the center. Mouse LECs with positive staining for α-SMA however did not show any sign of cell alteration [40]. LEC in their native state proliferate in a monolayer of cells. With cells seeded on top of the gels this did not happen. Instead LEC appeared to be present in clusters of cells producing α-SMA. Only LMWG BM and Matrigel were able to induce reorganization of the cytoskeleton and supported formation of a layer of cells, although the size of the nuclei increased and the cytoskeleton was more disorganized on Matrigel. Nagamoto et al. also described that in the early stages of EMT the α-SMA in LEC was expressed in a granular pattern, whereas in mature stages α-SMA positive stress fibers were clearly visible [9]. These different levels of differentiation of LEC were also detected with our different hydrogelators. For LMWG BM the quantity of α-SMA produced was low with the protein predominantly being located near the nucleus. In the presence of
connecting cells as fibers. This indicates the presence of premature and mature levels of mesenchymal differentiation of LEC seeded on LMWG BM and Matrigel, respectively. The LMWG IKVAV+YIGSR demonstrated a total alteration of LEC present as clusters or sometimes in cell layers with α-SMA mainly present as stress fibers. We conclude that LMWG BM appears to delay the EMT process of LEC when the cells are seeded on top of the hydrogel.
In the bottom model, no large differences in LEC metabolic activity were found among the different hydrogels. But the detection of α-SMA, the size of the nucleus and the expression of EMT-related genes were different between LEC grown underneath the different gels. Cells experienced an EMT-like response in the beginning of the culture with gels on top of them by expressing α-SMA in granular and fiber shapes, an increase in the size of nuclei and disorganization in the cytoskeleton. With prolonged culture these markers become less significant and obvious. At the end of the 5-day culture, a decrease in the nucleus size and in α-SMA expression was most evident in LEC underneath LMWG BM and LMWG IKVAV+YIGSR but not underneath Matrigel. These observations point to a better connection between LEC and peptides (both BM and IKVAV+YIGSR) and to a reduction in EMT markers in LEC underneath LMWG BM.
The different cell interactions in top and bottom models were also present at the mRNA level. Most of the fibrotic genes studied exhibited different patterns in LEC grown in top or bottom models, except for LMWG BM. This indicated that LEC had a favorable connection, without large signals of EMT differentiation, with this combination of peptides. In general, LEC on top of LMWG and LMWG IKVAV+YIGSR showed the highest expression of most of the fibrosis-related genes. From the α-SMA mRNA expression it was not possible to differentiate between “mature” or “early” α-SMA, however it was shown that all gels stimulated LEC to downregulate α-SMA mRNA compared with LEC cultured on Permanox. For the bottom model, LEC in contact with the LMWGs expressed lower amounts of fibrotic genes than in contact with Matrigel. This correlates with the immunohistochemistry results. In general, LMWG BM was demonstrated to slow down the development of EMT. In our study LEC were cultured in the presence of serum that by itself is an inducer for EMT [41]. Since the LMWG BM may be able to restore the characteristics of the basement membrane, it is possible that LEC reduced EMT because the cells are able to restore apical-basal polarity. Ours results suggest that the percentage and combination of peptides in the BM mix was more favorable to the inhibition of EMT than Matrigel which is also based on basement membrane proteins. But the combination of solely laminin-derived peptides IKVAV (0.5%) and YIGSR (0.5%) was not enough to decrease EMT in LEC. In fact, the native lens basement membrane includes a mix of laminin and type IV collagen. Fibronectin however is detected in large amounts during embryogenesis, but is a minor compound in the adult lenses. Olivero et al, showed that LEC proliferation and adhesion were influenced by both laminin and type IV collagen, whereas fibronectin was more important for cell differentiation and migration [42]. These observations can explain why the laminin mix (IKVAV and YIGSR) is able to transform clusters of LEC into a layer of cells but failed to maintain the epithelial phenotype. On the other hand, the addition of
3
morphology must be taken into consideration. For Matrigel the top model is the seeding geometry advised by the supplier. Therefore, the high values for Δt (difference between cell metabolic activity on day 5 and day 2) were expected. A low metabolic activity of LEC on top of LMWG (without peptides) was synonym for quiescent cells or apoptotic cells, present in cell clusters. This indicates that LMWG behaves as a repellent gel that does not readily support cell adhesion despite the presence of a three-dimensional nanofiber network. The addition of peptides to the LMWG yielded an increase in cell metabolic activity mainly for the mixture of peptides that resembles the basement membrane most, the LMWG BM. This was associated with a transition of cell clusters to a more natural-like monolayer of cells. The basal lamina is the natural surface for epithelial cell attachment [38]. However, when the cells were seeded underneath the gels, the metabolic activity ratio was independent of the gel composition with a clear cell layer formation from day 2 to day 5 of culture. This can be explained by the favorable interactions between Permanox and LEC when compared with LEC-gel interactions. In this case, the low metabolic activity cannot be referred to as stress or quiescence but rather as a passive stage where the cells have reached a comfortable cell attachment with corresponding low metabolic activity. Permanox was chosen for this study after several experiments with other cell culture bottoms such as type I collagen and fibronectin coatings and glass (data not shown). In all these cases LEC detached from these materials as an intact layer of cells after addition of the gel components.
Besides the increase of α-SMA during the ongoing EMT process, the shape of the cells also dramatically changes. During EMT the small and hexagonal epithelial cells transform into much larger myofibroblasts that are more similar to mesenchymal cells [8, 39]. When the EMT process has “matured” and cells end up in a fibrotic stage, cells will become more similar to fibroblasts. In general, different stages of EMT can be related with a different marker distribution. Nagamoto et al. reported that mammalian LEC start to express α-SMA almost immediately after culture of the primary cells, which by the way is usually done on stiff materials such as tissue culture plastics [9]. This was also visible in our control cultures where a slight increase of this protein could be observed. Garcia et al. mentioned that it is more appropriate to analyze α-SMA with both qualitative and quantitative tools rather than only detection of the protein. In that study LEC explants from bovine, rabbit and human lenses expressed α-SMA from the start of the culture in the equatorial zones of the lens but not in the center. Mouse LECs with positive staining for α-SMA however did not show any sign of cell alteration [40]. LEC in their native state proliferate in a monolayer of cells. With cells seeded on top of the gels this did not happen. Instead LEC appeared to be present in clusters of cells producing α-SMA. Only LMWG BM and Matrigel were able to induce reorganization of the cytoskeleton and supported formation of a layer of cells, although the size of the nuclei increased and the cytoskeleton was more disorganized on Matrigel. Nagamoto et al. also described that in the early stages of EMT the α-SMA in LEC was expressed in a granular pattern, whereas in mature stages α-SMA positive stress fibers were clearly visible [9]. These different levels of differentiation of LEC were also detected with our different hydrogelators. For LMWG BM the quantity of α-SMA produced was low with the protein predominantly being located near the nucleus. In the presence of
connecting cells as fibers. This indicates the presence of premature and mature levels of mesenchymal differentiation of LEC seeded on LMWG BM and Matrigel, respectively. The LMWG IKVAV+YIGSR demonstrated a total alteration of LEC present as clusters or sometimes in cell layers with α-SMA mainly present as stress fibers. We conclude that LMWG BM appears to delay the EMT process of LEC when the cells are seeded on top of the hydrogel.
In the bottom model, no large differences in LEC metabolic activity were found among the different hydrogels. But the detection of α-SMA, the size of the nucleus and the expression of EMT-related genes were different between LEC grown underneath the different gels. Cells experienced an EMT-like response in the beginning of the culture with gels on top of them by expressing α-SMA in granular and fiber shapes, an increase in the size of nuclei and disorganization in the cytoskeleton. With prolonged culture these markers become less significant and obvious. At the end of the 5-day culture, a decrease in the nucleus size and in α-SMA expression was most evident in LEC underneath LMWG BM and LMWG IKVAV+YIGSR but not underneath Matrigel. These observations point to a better connection between LEC and peptides (both BM and IKVAV+YIGSR) and to a reduction in EMT markers in LEC underneath LMWG BM.
The different cell interactions in top and bottom models were also present at the mRNA level. Most of the fibrotic genes studied exhibited different patterns in LEC grown in top or bottom models, except for LMWG BM. This indicated that LEC had a favorable connection, without large signals of EMT differentiation, with this combination of peptides. In general, LEC on top of LMWG and LMWG IKVAV+YIGSR showed the highest expression of most of the fibrosis-related genes. From the α-SMA mRNA expression it was not possible to differentiate between “mature” or “early” α-SMA, however it was shown that all gels stimulated LEC to downregulate α-SMA mRNA compared with LEC cultured on Permanox. For the bottom model, LEC in contact with the LMWGs expressed lower amounts of fibrotic genes than in contact with Matrigel. This correlates with the immunohistochemistry results. In general, LMWG BM was demonstrated to slow down the development of EMT. In our study LEC were cultured in the presence of serum that by itself is an inducer for EMT [41]. Since the LMWG BM may be able to restore the characteristics of the basement membrane, it is possible that LEC reduced EMT because the cells are able to restore apical-basal polarity. Ours results suggest that the percentage and combination of peptides in the BM mix was more favorable to the inhibition of EMT than Matrigel which is also based on basement membrane proteins. But the combination of solely laminin-derived peptides IKVAV (0.5%) and YIGSR (0.5%) was not enough to decrease EMT in LEC. In fact, the native lens basement membrane includes a mix of laminin and type IV collagen. Fibronectin however is detected in large amounts during embryogenesis, but is a minor compound in the adult lenses. Olivero et al, showed that LEC proliferation and adhesion were influenced by both laminin and type IV collagen, whereas fibronectin was more important for cell differentiation and migration [42]. These observations can explain why the laminin mix (IKVAV and YIGSR) is able to transform clusters of LEC into a layer of cells but failed to maintain the epithelial phenotype. On the other hand, the addition of
